US3453143A - Method of making a moisture sensitive capacitor - Google Patents

Description

y 1959 A. E. MEIXNER ET AL 3,453,143

METHOD OF MAKING A MOISTURE SENSITIVE CAPACITOR Original Filed Oct. 28. 1964 Sheet SU BSTRAT E m w lulL @Mm NIT S A S TULWUC WE E W N D f; K IE 4 AEO I O TTN /5 SSA 2 1 R E EH F T WA 11 E H Du G O E vH G KCG.A A U N L MVPMO I A Kl] CCV D m L m m R Em A Mmm G m mm HM S w ROS E R NWm N V AE w E E M W W m 5K1 R W R E TK R a CO HE E E5 D E RR D R AF E Y 5% B U P VK m P 4 O 3 0 2 C E B L H EM 0 O O 0 w 0 w 3 m 1 $301 FIG.2

ATTORNEYS y 1, 1969 A. E. MEIXNEh ET AL 3,453,143

METHOD OF MAKING A MOISTURE SENSITIVE CAPACITOR Original Filed Oct. 28. 1964 Sheet Z of '1 llllllHHIl IIHIHJHH lllHllllHl lllIlHHlH 10 I 10 10 ANGSTROMS/SEC FIG 3 11111 IN! 11mm HUR E. MEIXNER DERICK W. REYNOLDS INVENTORS BY M CL M ATTORNEYS .A. E. MEIXNER ETAL July 1, 1969 3,453,143

METHOD OF MAKING A MOISTURE SENSITIVE CAPACITOR Original Filed Oct; 28, 1964 Sheet 40 KV- 6 AMPS ELECTROSTATIC GUN SPACE CHARGE CURRENT LIMITED CATHODE C ENTER CURVATURE ARTHUR E. MEIXNER FREDERICK W. REYNOLDS INVENTORS ,4 Q.

ATTORNEYS July 1, 1969 I A. E. MEIXNER ET AL 3,453,143 METHOD OF MAKING A MOISTURE SENSITIVE CAPACITOR Original Filed Oct. 28. 1964 I snub-4 01 7 GOLD ELECTRODE PLATE D Si 0 DIELECTRIC I7//////W k\\\\\\\\/%% I GLASS SUBSTRATE Rp FIG. 5 FIG.

CAPACITANCE vs. FREQUENCY IN ROOM AIR CAPACITY C un IN DESICCATOR 1 I IIIIIII l Illllll FREQUENCY (CP. 5.)

7 ARTHUR E. MEIXNER I FRERICK W REYNOLDS INVENTORS BY j. a. WM

ATTORN S July 1, 1969 Original Filed Oct. 28, 1964 DISSIPATION FACTOR -D- DISSI A. E. MEIXNER ETAL 3,453,143

METHOD OF MAKING A MOISTURE SENSITIVE CAPACITOR Sheet 5 of? PATION FACTOR vs. FREQUENCY III'III IN ROOM AIR I N DESICCATOR lIllll l lIlllIl I 10 FREQUENCY (CPS) FIG. 8'

ARTHUR E. MEIXNER FREDERICK W. REYNOLDS INVENTO S ATTORNEYS July 1, 1969 A. E. MEIXNER ET AL 3,453,143

METHOD OF MAKING A MOISTURE SENSITIVE CAPACITOR Original Filed Oct. 28, 1964 Sheet 6 of 7 IHIIHIHL IIIIHIIIIH HHIHIIHI IIIIIHIHE E {SJESICCATOR g E IN 5 I ROOM AIR 1O4: moms) 5 g 10 E E J z RATIO 5 L I 1o E 5 i llllillllll mnmml mmmu \mmmfi' 1o 10 10 10 10 CYCLES/SEC.

FIG. 9

ARTHUR E. MEIXNER FREDERICK W REYNOLDS INVENTORS BY 4 film ATTORNEYS RESPONSE TO MOISTURE AMBIENT O-I-A u y 1, 1969- A. E. MEIXNER ETAL 3,453,143

METHOD OF MAKING MOISTURE SENSITIVE CAPACITOR v Original Filed Oct. 28, 1964 Sheet 7 of 7 EXPOSED To ROOM AIR o 5% LL] m ,r REPLACED I IN DESICCATOR I z I Z I i O 1 2 3 4 I TIME(MIN) FIG. 10b

7 ARTHUR E. MEIXNER .CONST I 63x10 AM FREDERICK w REYNOLDS INVENTORS FIG. 10a 0 I 7 BY A 7. mm aw ATTORNEYS United Patent US. Cl. 117-217 2 Claims ABSTRACT OF THE DISCLOSURE Method of making a thin film capacitor which is extremely sensitive to moisture and humidity. The capacitor includes a glass substrate, a first thin film gold electrode plate, a homogeneous silicon dioxide dielectric thin film over the first gold plate, a second gold thin film plate partly over the silicon dioxide dielectric thin plate and over the substrate.

This application is a division of application Ser. No. 406,987, filed Oct. 28, 1964 and now abandoned.

The present invention relates to the manufacture of a moisture responsive electrical device and more particularly to such a device which can provide a fast response and can be used in electrical circuitry such as a regulator or control arrangement.

Small effects of humidity upon the electrical properties of thin film capacitors have been previously observed. However, the electrical effect was small and a considerable change in ambient moisture was required to produce any appreciable change in capacitance. For this reason, such capacitors have not been used to any great extent for this purpose. Although attempts may have been made to provide moisture responsive capacitors, none, as far as we are aware, have ever been used successfully in a practical system.

It has now been discovered that a very fast response time, moisture controlled, in thin film variable impedance may be made from fast evaporated quartz film. The order of magnitude of the response is sutficient for the device to be used in a system and the response time is in the range of seconds.

In the production of thin films, the thin film basic material: gas ratio during evaporation plays a very important part indetermining the evaporated film characteristics. In general, the greater this ratio, the closer the approach to many of the bulk properties. The materiakgas ratio for present methods of evaporation are severely limited, rarely exceeding 100: 1. Most evaporated films are highly contaminated by the ambient residual gas during evaporation. It is common for such contamination to attain values of several percent.

Another object of the present invention is to provide an evaporation method in which the thickness of the deposited film is uniquely determined by the total energy supplied (joules). This offers a simple and reliable method for control of film thickness, particularly in the manufacture of capacitors.

A further object of the present invention is to provide a method which yields essentially complete evaporation of all the material acted upon by the energy source. This is important when evaporating alloys. The components of an alloy, in general, have diiferent vapor pressure temperature values and hence tend to fractionate during evaporation. The composition of the film, therefore, usually ditfers from that of the source material.

Another object of the present invention is to greatly reduce the total amount of radiant energy received by the film during evaporation from the evaporating source since heating of the film in this manner alters the film characteristics.

Yet, another object of the present invention is to provide a thin film capacitor extremely sensitive to moisture.

Other objects and advantages of the present invention will be apparent from the following detailed description thereof when taken in conjunction with the accompanying drawing, in which:

FIG. 1 is a schematic explanation of the electrostatically focussed electron gun used herein;

FIG. 2 shows a plot of the total beam energy as a function of time for the gun depicted in FIG. 1;

FIG. 3 graphically illustrates the material:gas ratio as a function of deposition rate, obtained by using the gun depicted in FIG 1;

FIG. 4 illustrates the design of the shape of the anode and cathode for the gun depicted in FIG. 1;

FIG. 5 explains schematically the equivalent circuit of the bridge used in some of the measurements described in Example I;

FIG. 6 illustrates the capacitor produced by the method I described in Example 1;

FIG. 7 shows a plot of capacitance as a function of frequency for a capacitor made in accordance with the present inventive concept;

FIG. 8 shows a plot of the dissipation factor as a function of frequency for a capacitor made in accordance with the present inventive concept;

FIG. 9 shows a plot of the impedance as a function of frequency for a capacitor made in accordance with the present inventive concept;

FIG. 10a explains schematically the equivalent circuit used for the measurements described in Example I; and,

FIG. 10b shows a plot of the response performance as a function of time at a constant frequency determined by calculating the change in impedance from measurements taken at different time intervals on the bridge illustrated schematically in FIG. 1011.

Generally speaking, the present invention contemplates a thin film capacitor which is extremely sensitive to moisture and humidity. This capacitor generally includes a glass substrate, a first thin film gold electrode plate over the substrate, a homogeneous silicon dioxide dielectric thin film over the first thin film gold electrode plate, and a second thin film gold electrode plate partly over said silicon dioxide dielectric thin film and over said substrate. The thin film silicon dioxide dielectric material is deposited over the gold electrode by means of an electro-- statically focussed electron gun. This type of gun is particularly suitable for evaporation of thin films where the presence of stray inhomogeneous matter cannot be tolerated. Another feature of this particular gun is that it has been designed for operation as space charge current limited. This means that the current has a known and fixed value for a given geometry and voltage provided that the cathode emission is equal or greater than that required to supply the designated current. With this type of operation the beam power is a known function of time. Consequently, the same evaporation conditions can be accurately repeated.

In the electron gun shown in the drawing, cathode and anode are sections of concentric spheres having a radii ratio of 2.0. The emitter is of the button-type indirectly heated by another electron gun of simple structure. The peak power of the main beam is of the order of 250 kw. A somewhat similar gun, but only this peak power, has been described by R. Thun and I. B. Ramsey. (Vacuum Symposium Transactions, American Vacuum Society, 1959, page 192.)

If a total pulse energy of at least 2000 joules is needed for the evaporation of the material selected, 40 kv. can be selected as the charging voltage, 6 ,uf. as the capacity, and 6 amps as the maximum beam current.

Since the gun has been designed to operate space charge current limited, the current I at any time t after start of the pulse is given by:

I=K(V) where V is voltage across condenser at time t.

For the gun constants described later,

The expression for voltage remaining on the capacitor versus elapsed time t in seconds is given by:

[er-car The corresponding equation for beam current is:

Using the foregoing equations, a plot of total beam energy versus time for the gun of FIG. 1 is shown in FIG. 2. This curve pertains to the 40 kv. gun operated by discharge of the 6 ,uf. condenser which has been charged to this voltage. About 2000 joules of energy is applied by the beam in 15 milliseconds.

Material: gas ratios versus deposition rate and ambient gas pressure are shown in FIG. 3. This method of evaporation yields material: gas ratios in excess of 10 :1 at ambient gas pressures of 15 mm.

To select the gun design constants for the foregoing electron gun, the following expression for beam current is first used.

For:

I 29.33 10*V sin (0/2) where:

0 is the half gun aperture V is the gun volts A is a constant determined by the ratio of cathode anode radii, R /R For R /R =2.0, A =.75. (See J. R. Pierce, Theory and Design of Electron Beams, D. Van Nostrand, 1954, page 177.)

Using the above, we find that for a beam current of 6 amps that 0:16".

Now assuming a cathode emitter button 1.6 cm. diameter then:

R =.8/sin 0:2.9 cm.

R 1.45 cm.

While the above gun will operate, the beam collection efi'iciency of the emitted electrons would be poor. This is undesirable because of the high emission requirements imposed on the cathode as well as undue heating of the anode. Spagenberg in his book Vacuum Tubes, pages 456-457, shows electrode shapes determined experimentally to overcome this problem. This information has been used to design the electrode shapes which are shown in FIG. 4. The curved shapes shown by dots have been approximated with straight lines to facilitate machining operations.

Since very large thermionic currents are required for ultra-fast evaporation, it is desirable to operate the cathode at as high a temperature as possible consistent with cathode life and tungsten evaporated. To this end it is essential that the cathode be heated only for the necessary time to produce the single high current pulse. The heating time for the following cathode has been calculated as indicated below.

Radius of cathode mm 8 Thickness mm 2 Mass gms :8 Heat content, '(T-300) joules 1.3 Total area sq.cm 4 Emissivity .33

Due to the method of mounting of the button in which small loops of 5 mil tungsten wire are employed to reduce loss by conduction, the conduction losses can be neglected. These are less than 10% of the radiation loss.

If 650 watts are used to heat the button:

In a well-designed ultra-high vacuum system the ultimate pressure is governed by the outgassing rate of the wall of the system and the pumping speed provided. It is common practice to make the walls of such systems stainless steel with the inside surfaces highly polished. The high polish reduces the effective surface area for gas adsorption and trapping. The rate of gas evolution from such surfaces in a vacuum depends in part upon the previous history of the surface. Pumping alone is not very effective in reducing the evolution rate; in fact, even after 10 or more hours the rate may be l0 micron liters per square cm. per second. If the system has a surface area of 5x10 sq. cm., and a pumping speed of 500 liters/ sec. is available, the vacuum attainable at this time would be only 10 mm. Consequently, it is accepted practice to bake-out such systems to 300-450 C. After such bakeout the evolution rate may be lO- micron liters per sq. cm. per sec. or less. If there are no leaks and we still have a pumping speed of 500 liters/sec. the vacuum should reach 10 mm.

Due to the large heat capacity of such systems, most of the time is consumed in heating and cooling. Cycle times of 16 hrs. are usual.

Now the major part of the above gas load is water vapor adsorbed on the wall surface. If the surface could be heated to these temperatures without materially raising the average wall temperature, the cycle time should be greatly reduced. Such heating can be accomplished by using energy pulses of large magnitude and very short duration, and, of course, a restricted number of pulses in a given length of time. The restrictions on the total number of pulses can, of course, be removed if the exterior of the walls is cooled.

Surface heating may be calculated using well-established principles of transient heat flow.

In this application we are interested in a one-dimensional flow across a wall thickness. The thickness is assumed semi-infinite, and conductivity (thermal) and specific heat values are assumed constant over the temperature range of interest.

Let:

0=wall temperature at distance X 0 =surface temperature at X =0 F=incident heat flexcal./ sq. cm./sec. K=thermal conductivity, c.g.s. units k=diffusive tendency, K/

P=density, c.g.s. units Equation D has been used to calculate three different values of surface temperature as a function of joules and time.

For the pulse times shorter than 1.0 millisecond the pulse energy required per sq. cm. to reach a surface temperature of 300 C. is less than 0.7 joule. For pulse times of .2 millisecond only 0.3 joule is required.

These pulse energies are within the capabilities of existing flash lamps such as are used for laser operation. (N. A. Kuebler and L. S. Nelson-Radiant Energies and Irradiances of Capacitor Lamps-Iournal Optical Society of America-vol. 51, pp. 1411-16, 1961 and Carslaw and JaegerConduction of Heat in Solids--Oxford Press.)

In the construction employed for this gun the target or evaporative material is located in essentially a field free space. If positive ions of any significant magnitude are generated during evaporation, there will be a tendency to reduce the space charge and hence make the focussed spot smaller. To overcome this difficulty, it may be necessary to provide for positive ion collection within the target holder. This can be done by introducing an electrode having a negative potential relative to the anode and target.

Heretofore, when rapid evaporation of materials was carried out, there was a spluttering of the substrate rendering the product useless. However, by means of the above-described gun, a controlled evaporation can be carried out on almost any material and the results obtained are repeatable.

EXAMPLE I A prepared quartz microscope slide was used as a substrate. An evaporated thin film of gold of the order of 1000 A. thickness was then deposited on the quartz substrate to serve as one electrode plate of the capacitor. Standard techniques were employed in the deposition of the gold electrode plate such as described by L. Holland Vacuum Deposition of Thin Film, John Wiley & Sons, New York, 1956, chapter VI.

A quartz film was evaporated by the pulsed electron gun previously described. Three pulses of approximately 40 milliseconds duration were employed at maximum values of beam voltage and current, respectively, of 12 kv. and 1.0 ampere. The substrate was outgassed in the vacuum prior to film deposition at a temperature of ap proximately 350 C. The vacuum was 10* mm. of Hg.

The calculated thickness of the deposited film, using the weight lost by the target and the known evaporation distance, was approximately 1700 A. By the use of approximately shaped masks the top and bottom evaporated gold electrodes formed 8 capacitors from this quartz film, each having an area of .0625 sq. cm.

After evaporation of the dielectric film, the substrate was transferred to another vacuum station where it was reheated prior to evaporation of a second top gold film, also of the order of 1000 A. thickness. The finished capacitor was stored in a desiccator for several days, at which time measurements were made both in the desiccator and exposed to room air. The capacitor is shown in FIG. 6.

The capacity measured in the desiccator of each small square (.0625 sq. cm. area) agreed with that calculated for a film of this thickness and assumed dielectric constant of approximately 4. However, when the sample was exposed to room air it rapidly increased from 1.8x 10* to 55 10 farads. This change was completely reversible upon returning the sample to the desiccator.

The measured change in capacitance as a result of humidity of the capacitors of Example I were then tabulated in Table I.

TAB LE I Series 09/ Parallel acitance, Dissi- Series Parallel capaci- Frequency 0., nano pation reactance, impedtance, O (f) farads factor D X5 ance, Zp nano tarads In room air:

86 15 18,500 18,660 84 500.-. 75 .34 4,240 4, 560 67 1,000.-.. 65 46 2, 450 2, 700 54 5,000 38 .85 840 1, 104 22 10,000--- 25 88 637 827 15 20,000... 15 78 540 686 9. 3 In desiccator:

The equivalent circuit is shown in FIG. 5 and, from FIG. 5,

R =X /D The very large response of this quartz film capacitor to changes in moisture content of the surrounding ambi ent indicate its usefulness for indicating and measuring changes in humidity, etc. Its very fast response and essentially complete reversibility are important properties for such use.

Various other methods may also be used to read out the changes in dielectric properties with moisture, such as the control of an oscillator frequency or simply as a variable impedance.

Capacitance, impedance and dissipation factors, as a function of frequency, were measured on a bridge whose equivalent circuit is shown in FIG. 5 as the capacitor C produced in accordance with the present inventive concept in parallel with a resistor R and plotted on FIGS. 7, 8 and 9, while the response time performance at a constant frequency of 200 c.p.s. was determined by calculating the change in impedance from bridge measurements taken at different time intervals. The bridge equivalent circuit for this measurement is shown in FIG. 10a and includes a 200 c.p.s. oscillator in series with a one million ohm resistor. The capacitor produced in accordance with the present inventive concept described herein is parallel with a vacuum tube voltmeter. The plotted results are shown in FIG. 10b.

7 EXAMPLE 11 Measurement of adsorbed water vapor-thermocouple Since heat energy is involved in adsorption and absorption, it is possible to determine the amount of gas adsorbed by the film. Since the rate of change is very rapid, it is possible to make measurements with good accuracy by noting the change in temperature of the substrate and film. The essential requirements are that the heat capacity of the substrate be as small as possible and that the change in ambient be rapid.

Mica was used as the substrate since it could be split into a thin sheet one mil thick. The temperature change was measured with an evaporated thin film thermocouple of aluminum and per-malloy. The latter was found to give a sensitivity of 30 ,uv./ C. The quartz film was evaporated on top of the thermocouple and also over another separate aluminum film to provide a test capacitor for checking the performance of the quartz film as a dielectric.

The capacity measurements on the complete unit at diiferent ambient conditions of condensable gas showed that the film was similar in properties to others previously made.

The completed unit was then placed in a small vacuum desiccator and evacuated. During pump-down, the thermocouple indicated that gas was being evolved from the film, and it became colder. After steady state had been reached and the temperature of the substrate returned to that of the surroundings, air was suddenly let into the system. The thermocouple indicated a rise in temperature of 6 C. Using the mass of the substrate, its specific heat and the temperature rise, it was calculated that 7.4 ,ug. of water was absorbed per sq. cm. of film. (The relative humidity of the room air was approximately 50%.

EXAMPLE III A large number of films of quartz, aluminum oxide and cerium oxide have also been evaporated on a variety of substrates. All of these show absorption effects similar in nature, but to a much lesser degree. Substrates used successfully were of rods of glass, polished fused quartz, mica, single crystal NaCl, fused alumina with an alkali free glass glaze from Corning Glass Co.

A number of condensable gases other than water vapor were also used and similar absorption behavior noted. In general the change in capacity increases with the dielectric constant of the vapor used. The magnitude of the effect is a fraction of the partial pressure of the ambient gas, the larger the pressure, the greater the pressure.

At room temperature the change in capacity becomes larger in going from carbon tetrachloride to water.

Effect Carbon tetrachloride. 2. 2 Small. Trichlorethylene- 3. 4

Methylene chloride- 9. 1 Large. Methyl alcohol 33 Water 79 Very large.

K=dielectric constant liquid.

5;- ambient pressure. A rubber bulb attached to a closed volume containing the sample provides adequate short pulses of pressure change to be readily observed and measured.

It is to be observed, therefore, that the present invention provides for a method of manufacturing a thin film capacitor comprising the steps of depositing a first t-hin film electrode plate on a non-conducting substrate, then, depositing under vacuum, a thin film dielectric material of homogeneous silicon dioxide and suboxides, over one portion of the electrode plate by discharging electron pulses of the order of 250 kw. on quartz material so as to evaporate the quartz into silicon dioxide and suboxides thereof onto the electrode plate-containing substrate, and, depositing a second electrode plate in face-to-face relation wit-h said first electrode plate over said dielectric material and said one portion of said first lectrode plate.

Although the present invention has been described in conjunction with preferred embodiments, it is to be understood that modifications and variations may be resorted to Without departing from the spirit and scope of the invention, as those skilled in the art will readily understand. Such modifications and variations are considered to be within the purview and scope of the invention and appended claims.

What is claimed is:

1. A method of manufacturing an extremely humiditysensitive, thin film capacitor, comprising the steps of:

(a) outgassing a non-conductive substrate in a vacuum at a temperature of about 350 C.

(b) depositing a first thin film gold electrode plate of the order of about 1000 A. on said substrate;

(c) depositing under vacuum, a thin film dielectric material of homogeneous silicon dioxide and suboxides over one portion of the electrode plate by discharging electron pulses of the order of 250 kw. on quartz material so as to evaporate the quartz into silicon dioxide and suboxides thereof onto the electrode plate-containing substrate; and,

(d) depositing a second gold electrode plate likewise of the order of about 1000 A., in face-to-face relation with said first electrode plate over said dielectrlic material and one portion of said first electrode p ate.

2. A method as claimed in claim 1, including the step of reheating the dielectric-containing substrate prior to the evaporation of the second electrode plate thereon.

References Cited UNITED STATES PATENTS 3,095,527 6/1963 Barnes et al 117-217 X 3,113,253 12/1963 Ishikawa et a1. 117217 X 3,274,025 9/1966 Ostis 117-217 3,298,864 1/1967 Maylandt 1l7217 X FOREIGN PATENTS 882,174 6/1953 Germany.

OTHER REFERENCES Yarwood et al.: Vakuum Technik, 11 Jahrgang, Heft 5, pages 149 and 150.

WILLIAM L. JARVIS, Primary Examiner.

US. Cl. X.R.

Images